JAANIS JUHANSON Impact of phytoremediation and bioaugmentation on the microbial

(1)

DISSERTATIONES TECHNOLOGIAE CIRCUMIECTORUM UNIVERSITATIS TARTUENSIS

10

(2)

(3)

DISSERTATIONES TECHNOLOGIAE CIRCUMIECTORUM UNIVERSITATIS TARTUENSIS

10

JAANIS JUHANSON Impact of phytoremediation and bioaugmentation on the microbial

community in oil shale chemical

industry solid waste

(4)

Department of Genetics, Institute of Molecular and Cell Biology, Faculty of Science and Technology, University of Tartu, Estonia.

Dissertation is accepted for the commencement of the degree of Doctor philo- sophiae (in environmental technology) at the University of Tartu on May 7^th, 2010 by the Scientific Council on Environmental Technology, Faculty of Scien- ce and Technology, University of Tartu.

Supervisor: Jaak Truu, PhD (Institute of Molecular and Cell Biology, University of Tartu).

Opponent: Kim Yrjälä, PhD (Department of Biological and Environmental Sciences, University of Helsinki).

Commencement: Room No. 204, 18 Ülikooli str., Tartu, on June 21 in 2010, at ^th 14.15.

Publication of this dissertation is granted by the Graduate School of Biomedi- cine and Biotechnology, University of Tartu.

ISSN 1736–3349

ISBN 978–9949–19–386–8 (trükis) ISBN 978–9949–19–387–5 (PDF)

Autoriõigus: Jaanis Juhanson, 2010 Tartu Ülikooli Kirjastus

www.tyk.ee Tellimus nr 275

(5)

5

LIST OF ORIGINAL PUBLICATIONS

I Truu J, Heinaru E, Vedler E, Juhanson J, Viirmäe M, Heinaru A:

Formation of microbial communities in oil shale chemical industry solid wastes during phytoremediation and bioaugmentation. In Bioremediation of Soils Contaminated with Aromatic Compounds, H. J. Heipieper (ed.).

2007, 76, p 57–66.

II Juhanson J, Truu J, Heinaru E, Heinaru A: Temporal dynamics of microbial community in soil during phytoremediation field experiment. J Environ Eng Landsc Manag 2007, 15, 4:213–220.

III Juhanson J, Truu J, Heinaru E, Heinaru A: Survival and catabolic performance of introduced Pseudomonas strains during phytoremediation and bioaugmentation field experiment. FEMS Microbiol Ecol 2009, 70, 3:446–455.

The articles are reprinted with the permission from the copyright owners and the publishers.

My contribution to articles of current dissertation is following:

Ref. I – participation in soil sampling, laboratory experiments and data analyses.

Ref. II – participation in soil sampling, laboratory experiments, data analyses and article preparation.

Ref. III – participation in soil sampling, laboratory experiments, data analyses and article preparation.

(8)

ABBREVIATIONS

PAH Polyaromatic hydrocarbon MTBE Methyl tert-butyl ether TCE Trichloroethylene PCB Polychlorinated biphenyl 2,4-D 2,4-Dichlorophenoxyacetic acid PCP Pentachlorophenol

BTEX Acronym that stands for benzene, toluene, ethylbenzene, and xylenes

TNT 2,4,6-trinitrotoluene

RDX Hexahydro-1,3,5-trinitro-1,3,5-triazene HMX Octahydro-1,3,5,7-tetranitro-1,3,5-tetrazocine GC Gas-liquid chromatography

GC-MS Gas chromatography-mass spectrometry

HPLC High-performance liquid chromatography/High pressure liquid chromatography

LC-MS Liquid chromatography-mass spectrometry

D/TGGE Temperature/Denaturing gradient gel electrophoresis (T-)RFPL (Terminal) restriction fragment length polymorphism SSCP Single-strand conformation polymorphism

SSU rRNA Small subunit ribosomal ribonucleic acid CFU Colony forming unit

(Q)PCR (Quantitative) polymerase chain reaction AHB Aerobic heterotrophic bacteria

PDB Phenol degrading bacteria PCA Principal component analyses

(9)

9

1. INTRODUCTION

Oil shale mining, chemical processing (thermal treatment) and energetic use produce majority of solid waste in Estonia. Mining of oil shale in Estonia started in 1916 and reached its peak in 1980 when 31 million tons of oil shale were excavated per year (Kattai and Lokk, 1998). Nowadays the government allows to excavate up to 20 million tons of oil shale annually. The mining and processing of 1,000 million tons of oil shale up to now has been accompanied by the deposition of about 400 million tons of solid waste: more than 90 million tons of mining waste, ca 100 million tons of semi-coke (oil and chemical industry waste) and 200 million tons of combustion ashes (power generation waste). Currently ca 1.7 million tons of oil shale is treated thermally annually producing more than 200,000 tons of oil and oil shale chemicals. In addition, approximately 700,000 tons of semi-coke solid waste is disposed every year. As a result of oil shale thermal treatment during 85 years, semi-coke mounds cover an area about 180–200 ha in the north-eastern part of Estonia. Although semi- coke is produced in much smaller amounts than combustion ashes, semi-coke mounds are the most severe environmental concern in Estonia since semi-coke consists of high amount of different form of organic carbon that may pose hazard to the environment due to leaching of toxic compounds to both the surface water and the underlying aquifers, as well as the possibility of self- ignition. In addition, depositories have been historically used for dumping different wastes (e.g. oil pitch, waste sludge) therefore leachates from the depository area contain high concentration of oil products, phenol, cresols, dimethylphenols and resorcinols (Truu et al., 2002). According to the chemical properties and ecotoxicological tests fresh semi-coke is classified as hazardous waste while several years old semi-coke is practically neutral and considerably less toxic due to being rain-washed (Kahru and Põllumaa, 2006).

The storage of semi-coke in open dumps, as the present situation has developed, is not in accordance with European Union regulations and these land- fills must be closed by 2013. A standard approach to cover the landfill involves the utilization of different substrate layers like isolation, drainage and cover layer. As a result, the diffusion of pollution to the ground and surface waters and air is prevented. The problem with the standard solution is that only the isolation of the source of pollution is performed without further remediation of the pollutants. The number of alternative solutions is however limited due to the strict regulations on toxic wastes. Since only fresh semi-coke solid waste is classified as toxic waste, different management options for remediation can actually be considered for certain parts of the depository to truly achieve the degradation of the contaminants. Because of the enormous amount of the solid waste, the only feasible ways for the remediation include different technologies, which can be performed on site or with minimal need to move the waste.

A variety of remediation technologies are available for on-site remediation.

Soil vapour extraction, landfarming, bioventing, thermal desorption, biopiles are in situ remediation technologies, which have been used as real life appli-

3

(10)

cations for soil (reviewed by Khan et al., 2004). However, no single technology is appropriate for all contaminant types and the variety of site-specific conditions which exist at different contaminated sites. Site conditions, contaminant type, contaminant source, source control measures, and the potential impact of the possible remedial measure determine the choice of a remediation strategy and technology. Often more than one remediation technology is needed to effectively address contaminated site problems (Khan et al., 2004). In recent decades, phytoremediation as a cost effective and environmentally friendly technology has been used successfully for the remediation of soils contaminated with various pollutants.

(11)

11

2. THE AIM OF THE STUDY

The general aim of this thesis is to assess the feasibility of combination of phytoremediation and bioaugmentation as an alternative option for the remediation of semi-coke solid waste. Together with the use of vegetation, our purpose was to determine and target the kind of microorganisms in sampled habi- tats which are likely to suit specific conditions and remedial requirements.

Among those temporally and spatially prevalent microbial populations, strains which can degrade a specific contaminant are selected and identified for the development of a collection of biodegradative microbes that can be used together with vegetation and fertilizers for the remediation of oil shale solid wastes (Figure 1). As a result of the remediation approach based on simulta- neous use of plants (phytoremediation) and specific consortium of bacteria (bioaugmentation) concentration of pollutants will be reduced and semi-coke heaps will be covered with vegetation that prevents soil erosion and decreases amount of leachate.

Figure 1. Principal scheme of technological approach for remediation of oil shale chemical industry solid waste dump area.

The specific aims were:

• to study the impact of vegetation and introduced bacterial strains on the concentrations of pollutants as well as on the metabolic activity of indigenous microbial community in soil.

• to study the impact of vegetation and introduced bacterial strains on the dynamics of bacterial numbers and bacterial community composition in soil.

• to study the survival and catabolic performance of the introduced bacterial strains in soil for a prolonged period.

(12)

3. LITERATURE REVIEW 3.1. Phytoremediation

According to Cunningham and Berti (1993) phytoremediation is defined as the use of green plants to remove, contain, or render harmless environmental contaminants. In this process specially selected or genetically engineered plants are used which are capable of direct uptake of pollutants from the environment (Macek et al., 2000). Generally, phytoremediation of contaminants by a plant involves steps like uptake, translocation, transformation, compartmentalization, and sometimes mineralization (Schnoor et al., 1995). Factors affecting the uptake, distribution and transformation of organic compounds by a plant are mainly related to the physical and chemical properties of the compound (water solubility, molecular weight, octanol-water partition coefficient), as well as environmental conditions (e.g. temperature, pH, organic matter, and soil moisture content) and plant characteristics (e.g. root system, enzymes) (Suresh and Ravishankar, 2004; Susarla et al., 2002). Phytoremediation can be applied to both inorganic and organic pollutants present in solid and liquid substrate (Salt et al., 1998). Although the designations of different phytoremediation strategies vary in literature, the principal scheme is given in Figure 2.

Inorganic contaminants (heavy metals and radionuclides) can be either taken up from the soil and immobilized by the roots (phytoimmobilization), or transported to the plant shoot (phytoextraction) (Reichenauer and Germida, 2008). Since under most circumstances there is rather low bioavailability of metals in soil, including some metals that are essential to life, plants possess highly effective metal uptake system using transporter molecules such as zinc- regulated transporter protein, copper transporter protein etc. (Krämer et al., 2007). In addition, plants are capable of secreting metal-chelating molecules like siderophores and organic acids (malate, citrate), and biosurfactants such as rhamnolipids to the surrounding soil, but also extruding protons from the roots to acidify the soil and mobilize soil bound metals (Fig. 2) (Eapen and D'Souza, 2005; Garbisu and Alkorta, 2001). Inside the plant, heavy metals cannot be biodegraded but are only transformed from one oxidation state or organic complex to another (Garbisu and Alkorta, 2001). As a result, metals tend to accumulate in the plant. Nearly 450 hyperaccumulator plants ranging from annual herbs to perennial shrubs and trees (e.g. tobacco, sunflower, mustard, maize, pennycress, brake fern, Russian thistle, rattlebush, python tree, willow, poplar) have been described to accumulate and detoxify extraordinary high levels of metal ions, such as Ni, Co, Pb, Zn, Mn, Cd, etc. in their above ground tissues (Meagher, 2000; Padmavathiamma and Li, 2007; Shah and Nongkynrih, 2007; Sheoran et al., 2009). It has been suggested, that the prevention of herbivory and disease may be the main function of hyperaccumulation for the plant (Fattorini et al., 2010; Shah and Nongkynrih, 2007). Still, in this case it is possible to harvest and remove plants from the site after remediation for disposal or recovery of the contaminants (Susarla et al., 2002). For some

(13)

13

inorganic elements (Hg, As, Se) uptake by roots followed by transport to the shoot and transpiration to the atmosphere through the leaf stomata (phyto- volatilization) have been observed (Padmavathiamma and Li, 2007; Pilon- Smits, 2005). Since the volatile forms of Hg and Se are also toxic, it is question- able whether the volatilization of these elements into the atmosphere is desirable or safe (Padmavathiamma and Li, 2007; Watanabe, 1997).

Organic pollutants in soil like chlorinated solvents and polyaromatic hydrocarbons (PAHs) can be taken up and immobilized by plant roots (Gao and Zhu, 2004) as well as transpired from the shoot (methyl tert-butyl ether – MTBE, trichloroethylene – TCE, ethyl-benzene, xylene) (Ma and Burken, 2003; Ma et al., 2004). In addition, plants are capable of metabolizing organic contaminants (phytodegradation). The metabolism of contaminants by a plant can be di- vided into three phases: transformation, conjugation and compartmentalization (Fig. 2). In the transformation phase, contaminant is chemically modified (oxidation, reduction, and hydrolysis) and transformed into more polar, water soluble form by enzymes such as cytochrome P450 or carboxylesterases. By conjugation with endogenous molecules like sugars or peptides, the transformed contaminant is made less phytotoxic by glycosyltransferases and glutathione S- transferases, followed by compartmentalization phase where contaminant is transferred to the various compartments of the cell (storage in the vacuole or integration into cell wall) or in some cases excreted from the cell (Eapen et al., 2007; Ma and Burken, 2003; Reichenauer and Germida, 2008). However, there is a principle difference between metabolism of contaminants by a plant and by microorganisms – most contaminants are not utilized as a source of C, N and energy by plant since plants do not possess complete catabolic pathways for degradation and mineralization of pollutants (Eapen et al., 2007; Schroder and Collins, 2002). Frequently, during the degradation process even more toxic by- products (from the human point of view) may be produced compared to the initial pollutant. For instance, the transformation of TCE into trichloroethane, or the release of some metabolites from volatile pollutants into the environment by evapotranspiration have been detected (Burken and Schnoor, 1998; Ma and Burken, 2003). Only a few contaminants, for example PCBs, PAHs, nitro- aromatics and linear halogenated hydrocarbons can be completely mineralized by plants such as poplar, willow, alfalfa and different grass varieties (Kuiper et al., 2004; Meagher, 2000).

4

(14)

Figure 2. Phytoremediation of various organic and inorganic pollutants in soil. Plants are capable of removing organic and inorganic contaminants from soil by roots (phyto- immobilization), but also transporting and concentrating them in the harvestable part of the plant (phytoextraction/accumulation). In some cases transpiration to the atmosphere through leaf stomata may follow (phytovolatilization). Organic contaminants can be metabolized inside the plant (phytodegradation) in three sequential steps (phase 1 – transformation, phase 2 – conjugation, phase 3 – compartmentalization) using enzymes, such as CYP450 – cytochrome P450; GT – glycosyltransferase, resulting in the storage of the contaminant in the vacuole, integration into the cell wall, or excretion from the cell. In addition, organic contaminants can be degraded by plant-associated microorganisms in the rhizosphere (rhizodegradation). Plants facilitate the biodegradation of contaminants by releasing root exudates and other compounds to the surrounding soil as well as providing surface for the colonization of microbes, contributing in this way to the increased number and metabolic activity of microorganisms (rhizosphere effect) and enhanced bioavailability of the contaminant. Figure 2 is modified from Gerhardt et al., 2009; Reichenauer and Germida, 2008; Sheoran et al., 2009.

(15)

15

Transgenic plants can be developed by transferring genes from organisms which have the potential for degradation/mineralization of xenobiotic pollutants to candidate plants to improve the ability of plants to degrade/metabolize xenobiotic pollutants. Genes involved in the degradation of xenobiotic pollutants can be isolated from bacteria/fungi/animals/plants and introduced into candidate plants using Agrobacterium mediated or direct DNA methods of gene transfer (Eapen et al., 2007). Specific catabolic genes essential for the degradation of a contaminant are overexpressed in a plant, resulting in enhanced phytoremediation. For example, transgenic tobacco, arabidopsis, mustard, poplar, rice, potato have been reported to be able to improve phytoextraction, phytovolatilization and phytodegradation of heavy metals and organic contaminants like explosives, chlorinated solvents, PAHs, polychlorinated biphenols, various herbicides, and atrazine (Cherian and Oliveira, 2005; Eapen et al., 2007;

Rylott and Bruce, 2009). The most recent and very promising approach to improve phytoremediation ability is the construction of plants with enhanced secretion of enzymes capable of degrading xenobiotics into the rhizosphere (Abhilash et al., 2009; Gerhardt et al., 2009). The advantage of this method is that the plants do not need to take up the pollutants in order to detoxify them;

instead, the secreted enzymes can degrade the pollutants in the rhizospheric zone (Kawahigashi, 2009). However, there are strict regulatory restrictions for in situ applications of genetically modified organisms in the European Union and promising results have been obtained only in the laboratory and greenhouse experiments.

3.1.1. Rhizosphere effect and rhizodegradation

In addition to the plant metabolic capacity for the degradation of contaminants, plants have another important function in phytoremediation – plant roots establish favourable conditions for the microbes in rhizosphere, in this way facilitating the biodegradation of the contaminants (rhizodegradation). The term “rhizosphere” describes the portion of soil where microorganism-mediated processes are under the influence of the root system. In comparison to the bulk soil, increased metabolic activity, number and in certain cases, phylogenetic diversity of species of microorganisms can be found on the surface and in the vicinity of roots (Kuiper et al., 2004). This “rhizosphere effect” is mainly caused by the chemical impact (root exudates), supported by the physical (i.e.

gas exchange, soil moisture) impact of plant roots on the soil (Fig. 2) (McCully, 1999). Root exudates contain organic acids (lactate, acetate, oxalate, succinate, fumarate, malate, and citrate), sugars and amino acids as main components but also secondary metabolites (isoprenoids, alkaloids, and flavonoids) which are released to the soil as rhizodeposits (Jones, 1998; Singer et al., 2003; Singh et al., 2004). The amount and composition of root exudates is specific to plant family or species. It has been suggested that 10–44% of the photosynthetically fixed carbon is excreted by rhizodeposition (Bais et al., 2006; Kumar et al.,

(16)

2006). As a result, nutrient rich environment in the vicinity of roots is created.

Root exudates can be used as an energy source by microorganisms but since the structure of many secondary metabolites resembles those of contaminants, root exudates can also induce the expression of specific catabolic genes in microorganisms necessary for the degradation of the contaminant (Singer et al., 2003). This is important under many circumstances where microorganism cannot rely on energy gain from the contaminant and cometabolism is the only route for the degradation of contaminant. For instance, plant secondary meta- bolite salicylate has been linked to the microbial degradation of PAHs (naphthalene, fluoranthene, pyrene, chrysene) and PCB (Chen and Aitken, 1999; Master and Mohn, 2001; Singer et al., 2000), while terpenes can induce the microbial degradation of toluene, phenol, TCE (Kim et al., 2002). Therefore, roots can regulate the soil microbial community by producing and releasing root exudates and exoenzymes to the surrounding soil as well as providing surface for the colonization of microbes resulting in enhanced biodegradation of the contaminants. In turn, rhizosphere microorganisms provide plant with nitrogen, phosphorus and other minerals through decomposition of soil organic matter (Fig. 2). Rhizodegradation applies to a wide range of contaminants including those which due to their physicochemical properties are taken up by plants only in very small amount (higher-ring PAHs as an example; Reichenauer and Ger- mida, 2008). In many cases, rhizosphere microbes are the main contributors to the degradation process.

A recent strategy to improve phytoremediation and detoxification of contaminants is the use of endophytic bacteria. Endophytic bacteria are described as non-pathogenic bacteria and they seem to have a ubiquitous existence in most if not all higher plant species. They often belong to genera commonly found in soil, including Pseudomonas, Burkholderia, Bacillus and Azospirillum (Lodewyckx et al., 2002; Moore et al., 2006; Yrjala et al., 2010). Endophytic bacteria are also known to have plant growth promoting and pathogen control activities (Berg et al., 2005; Ryan et al., 2008). A major advantage of using endophytic bacteria over rhizospheric bacteria in phytoremediation is that while a rhizospheric bacterial population is difficult to control, and competition between rhizospheric bacterial strains often reduces the number of the desired strains (unless metabolism of the pollutant is selective), the use of endophytes that naturally inhabit the internal tissues of plants reduces the problem of competition between bacterial strains (Doty, 2008; McGuinness and Dowling, 2009). Studies suggest that these bacteria can be used to complement the metabolic potential of their host plant through direct degradation (Barac et al., 2004; Germaine et al., 2006; Phillips et al., 2008; Phillips et al., 2009) as well as transfer of degradative plasmids to other endophytes (Taghavi et al., 2005;

Wang et al., 2007). To date, many successful cases of phytoremediation of various contaminants using rhizospheric or endophytic bacteria have been reported (Table 1).

(17)

Table 1. Examples of successful phytoremediation cases of different contaminants using rhizospheric bacteria (RH) or endophytic bacteria (EN). Microbes PlantsContaminant Strategy Reference Pseudomonas sp. Sugar beetPCBs RH (Villacieros et al., 2005) RockcressPCBs RH (Narasimhan et al., 2003) Alfalfa PCBs RH (Brazil et al., 1995) Wheat TCE RH (Yee et al., 1998) Wild rye Chlorobenzoic acid EN (Siciliano et al., 1998) Pea 2,4-D EN (Germaine et al., 2006) Poplar MTBE, TCE, BTEX EN (Germaine et al., 2004; Moore et al., 2006) Pea Naphthalene EN (Germaine et al., 2009) Barmultra grass Naphthalene RH (Kuiper et al., 2001; Kuiper et al., 2004) Barley Phenanthrene RH (Anokhina et al., 2004) Sinorhizobium meliloti P221 Common reed Phenanthrene RH (Golubev et al., 2009) Indigenous degradersSwitchgrass PCBRH (Chekol et al., 2004) Red clover; Ryegrass 2,4-D RH (Shaw and Burns, 2004) Ryegrass PCP RH (He et al., 2005) White mustardPetroleum hydrocarbonsRH (Liste and Prutz, 2006) Hybrid poplarBTEX, toluene RH, EN(Barac et al., 2009) English oak; common ashTCE, toluene RH, EN(Weyens et al., 2009) BirchPAHs RH (Sipila et al., 2008) Altai wild rye; tall wheat grass Petroleum hydrocarbonsEN (Phillips et al., 2009) Burkholderia cepaciaYellow lupineToluene EN (Barac et al., 2004) Poplar Toluene EN (Taghavi et al., 2005) Barley 2,4-D RH (Jacobsen, 1997) Azospirillum lipoferum spp. WheatCrude oil RH(Muratova et al., 2005; Shaw and Burns, 2004) Azospirillum brasilense Cd; Enterobacter cloacae CAL2; P.putida UW3 Tall fescue grass PAHsRH (Huang et al., 2004) Methylobacterium populi BJ001PoplarTNT, RDX, HMXEN(Van Aken et al., 2004a; Van Aken et al., 2004b) Herbaspirillum sp. K1Wheat PCBs, TCP EN (Mannisto et al., 2001)

5

(18)

To summarise, phytoremediation is technology that is based on the combined action of plants and their associated microbial communities to degrade, remove, transform, or immobilize toxic compounds located in soils, sediments, ground water and surface water. Phytoremediation has been used to treat many classes of contaminants including petroleum hydrocarbons, chlorinated solvents, pesticides, explosives, heavy metals and radionuclides, and landfill leachates (Joner and Leyval, 2003; Susarla et al., 2002). There are several advantages of phytoremediation compared to conventional techniques, such as low cost, low disruptiveness to the environment, public acceptance, and potentiality to reme- diate various pollutants (Macek et al., 2000; Pilon-Smits, 2005; Susarla et al., 2002). In addition, plants as autotrophic systems with large biomass require only a modest nutrient input, and they also prevent the spread of contaminants through water and wind erosion (Cherian and Oliveira, 2005). Candidate plant for phytoremediation should have the characteristics such as high biomass production, extensive root system, and ability to tolerate high concentration of pollutant and withstand environmental stress. Like other methods, phytoremediation has its disadvantages e.g. climatic and geological limitations, potential phytotoxicity of contaminant, potential for the contaminant or its metabolites to enter the food chain, and potentially longer timescale compared to other technologies (Macek et al., 2000). Although some success has been reported using plants alone in bioremediation (Gerhardt et al., 2009; Pilon- Smits, 2005), the use of plants in conjunction with plant associated (rhizosphere or endophytic) bacteria offers more potential for bioremediation (McGuinness and Dowling, 2009).

3.2. Bioaugmentation

As pointed out above, plants frequently lack the metabolic capacity for the degradation of many contaminants. Unlike plants, microorganisms are commonly able to degrade and mineralize vast variety of different pollutants. The problem is that usually we deal with the pollutant which is a very complex compound or is a mixture of different compounds and is therefore degradable or mineralizable by very specific set of microorganisms (consortium). It is suggested that several microbial populations together degrade pollutants more efficiently than a single strain due to the presence of partners which use the various intermediates of the degradation pathway more efficiently (Heinaru et al., 2005; Pelz et al., 1999). During rhizodegradation, the degradation of a pollutant, in many cases, is also the result of the action of a microbial consortium. But even when the appropriate catabolic traits are present in the local microbial community, the abundance and activity of the microorganisms at the site may be too low for successful bioremediation. In this case bioaugmentation strategy is used. Bioaugmentation is a method to improve degradation and enhance the transformation rate of xenobiotics by the inoculation of specific microbes, able to degrade the xenobiotics of interest. Introduction of one single

(19)

19

strain with the complete degradation pathway as well as a consortium of bacteria, each with different parts of the catabolic degradation route involved in the degradation of a certain pollutant/intermediate can be applied for bioaugmentation (Kuiper et al., 2004). Bioaugmentation is performed either directly by introduced microorganisms, or via the transfer of catabolic genes to the local microbial community (plasmid mediated bioaugmentation). Regardless of the approach chosen, the isolation and characterization of the appropriate microorganisms as well as their survival and catabolic activity in the contaminated environment are the key factors for successful bioaugmentation (Thompson et al., 2005). According to Dejonghe et al. (2001), bioaugmentation should aim for the rearrangement of the group of organisms dominantly involved in the overall energy flux, so that specific catabolic traits necessary for the cleanup of pollutants are part of that active group.

The advantage of bioaugmentation relies on its ability to accelerate the removal rate of pollutant several fold over a relatively short time scale. How- ever, studies frequently observe that improvement of the bioremediation activity is temporary and the number of exogenous microorganisms decreases shortly after the addition of the biomass to the site (Bouchez et al., 2000a; Bouchez et al., 2000b; Ruberto et al., 2003). Several possible reasons for the failure of inocula, such as abiotic (extremes in temperature, water content, pH, nutrient availability, low availability or potentially toxic level on pollutants) and biotic factors (antibiotic production, antagonistic interactions) have been reported (Bouchez et al., 2000a; Gentry et al., 2004). According to Thompson and coauthors (2005), the most important factor determining the result of bioaugmentation comes first of all from the initial strain selection step. It has been suggested that the best way to increase the survival of the inoculum is to look for candidate microorganisms from the same ecological niche as the polluted area (El Fantroussi and Agathos, 2005). Such microorganisms are more adapted to the biotic and abiotic conditions in the polluted environments. It is also easier to incorporate them to the local microbial community. In addition, factors like inoculum density, physiological state and modes of introduction are known to considerably influence the survival and performance of the introduced microorganisms (Cunliffe et al., 2006; Thompson et al., 2005). Despite of the diffi- culties, successful bioaugmentation has been applied to both nonvegetated and vegetated soils (Jacques et al., 2008; Jézéquel and Lebeau, 2008; Ruberto et al., 2003; Siciliano et al., 1998; Singer et al., 2003), as well as activated sludge bio- reactor systems (Boon et al., 2000; Cordova-Rosa et al., 2009) treating contaminants, such as phenols, clorophenols, pesticides and oil products. Also, bio- stimulation approach based on the addition of nutrients (carbon, nitrogen, phosphorus, potassium) or electron acceptor/donors (acetate, nitrate, sulfate, glutamate) can be used in combination with bioaugmentation to improve the survival and catabolic activity of introduced microorganisms.

Plasmid-mediated bioaugmentation has been suggested as an alternative strategy where the survival of the introduced donor strain is no longer needed once catabolic genes are transferred and expressed in indigenous bacteria (Top

(20)

et al., 2002). It is known that genetic information encoding the degradation of xenobiotic compounds is often found on plasmids or other mobile elements, and this genetic information can be potentially transferred to the local microbial community also from the dead inoculum. Therefore, the idea behind this strategy is that more important than the survival of introduced bacteria is the survival of their catabolic traits. Successful plasmid-mediated bioaugmentation of organic contaminants and transfer of catabolic genes to the indigenous bacteria has been demonstrated in soils and in activated sludge systems (Bathe et al., 2004; Bathe et al., 2005; Mohan et al., 2009; Top et al., 1999; Top et al., 1998). Pseudomonas sp., Alcaligenes sp., Achromobacter sp., Comamonas sp., Burkholderia sp., Ralstonia sp., have been used as donor strains for the transfer of catabolic genes. Plasmid-mediated bioaugmentation has been suggested particularly in the context of rhizoremediation, as the rhizosphere may be a habitat that allows a higher frequency of catabolic gene transfer as well as higher metabolic activity compared with bulk soils, both of which are necessary for a successful plasmid-mediated bioaugmentation (Top et al., 2002).

3.3. Microbial community structure in the vicinity of roots

A multitude of biotic and abiotic factors, for example climate and season, grazers and animals, pesticide treatment, soil type, structure and history, plant health and developmental stage are assumed to influence the structural and functional diversity of bacterial communities in the rhizosphere (Garbeva et al., 2004; Germida and Siciliano, 2001; Graner et al., 2003; Jousset et al., 2006;

Siciliano et al., 2001). However, plant species (root morphology, plant age and health) and soil type (soil as the main reservoir for rhizosphere microorganisms) are considered to be factors that most substantially influence the structure and function of rhizosphere associated microbial community (Berg and Smalla, 2009; Kuiper et al., 2004). Although root exudates and the response of microorganisms to the latter as well as root morphology shapes the rhizosphere microbial communities (Berg and Smalla, 2009), it is not conclusively known whether plants are capable of actively select beneficial soil microbial communities in their rhizosphere through rhizodeposition process. It is also suggested, that there is actually no convincing evidence for direct stimulation of degrader microorganisms by plant roots through signalling, since increased abundance and activities of degrader populations in the rhizosphere could not be separated from other ecological interactions such as the effect of contamination (Siciliano and Germida, 1998; Wenzel, 2009). At least in theory, microbial strains or populations with degradation capabilities could be selected by a plant by microbial induced root exudation of compounds that can only be utilized by selected microorganisms (Dzantor, 2007; Siciliano and Germida, 1998). The evolutio- nary significance of a plant nourishing microbes in the rhizosphere is at least partially based on the protective value of the microbes in the root zone (Arthur

(21)

21

et al., 2005; Siciliano and Germida, 1998). Some data also suggest that beneficial symbioses and protective associations for the plant are encouraged in the rhizosphere, ensuring the supply of vital nutrients and changing the chemical and physical properties of the soil (Bais et al., 2004; Singh et al., 2004). Many microbes isolated from the rhizosphere are described to have plant/root growth- stimulating or growth-inhibiting properties. Beneficial microorganisms can benefit plant health by phytostimulatory and biofertilizing properties, inducing systemic response in the plant, resulting in the activation of plant defense mechanisms against various types of phytopathogens, and in terms of phytotoxicity of the contaminant, degrading the contaminant before it can negatively impact the plant (Chaudhry et al., 2005; Dams et al., 2007; Hontzeas et al., 2004; Liu et al., 2007; Raaijmakers et al., 2009). Also, the role of root-associated microbes in maintaining soil structure (i.e. aggregate stability) has been documented (Sen, 2003). Many of these beneficial rhizobacteria (mostly Pseudomonas spp.) can act as plant growth promoters as well as contaminant degraders (Cherian and Oliveira, 2005; Hontzeas et al., 2004; Kuiper et al., 2004) and this rhizodegradation efficiency may be a factor determining the selection of appropriate plant-contaminant degrader microorganisms pairs (Siciliano and Germida, 1998). Some findings suggest that instead of increasing the overall number of microorganisms, plants indeed select for taxonomic and functional groups in the rhizosphere regions (Bremer et al., 2007; Briones et al., 2002; Grayston et al., 2001), which are necessary for the degradation of specific contaminant (Leigh et al., 2006; Phillips et al., 2009; Siciliano et al., 2003).

3.4. Monitoring bioremediation

The majority of in situ bioremediation experiments have attempted to explain the efficiency of the process only on the basis of the kinetics of pollutant removal using direct (GC, GC-MS, HPLC, LC-MS, ion chromatography, proton nuclear magnetic resonance) or indirect (growth response of the pollutant degrading strain, appearance of degradation metabolites, consumption of O₂, or evolution of CO₂) methods (Baroja et al., 2005; Cledera-Castro et al., 2004;

Combourieu et al., 2004; Esteve-Nunez et al., 2005; Gea et al., 2004; Koren- kova et al., 2006; Pandey et al., 2009; Pieper et al., 2002). However, the diffe- rentiating metabolic degradation of the pollutant from the nonbiological removal is complicated and effective monitoring of microbial degradation under in situ conditions is rather poor because in many cases the decrease in the pollutant concentration may be observed as an outcome of the adsorbance of the pollutant to the environmental matrix (Pandey et al., 2009).

Currently it is generally acknowledged that in addition to the monitoring of pollutant removal, the environmental fate of degradative organism (i.e. the survival and activity) has to be monitored as well to maximize sustained bioremediation under natural conditions. Several cultivation based (microbial enumeration, soil enzyme activity analysis) and cultivation independent methods are

6

(22)

used for this purpose. However, because of the limitations of culture-dependent methods (for example the well known fact, that less than 1% of the microorganisms are culturable in standard laboratory conditions), molecular methods are mostly used nowadays for the study of microbial community structure and function. It is suggested, that DNA (extracted directly from the soil) based molecular techniques permit the most detailed determination of microbial community structural diversity and evaluation of the presence of certain functional genes in soil (Little et al., 2008). The majority of these techniques have been based on the sequencing/fingerprinting analysis of some phylogenetically relevant genes (such as 16S rRNA gene) amplified from the total community DNA. Among the most common fingerprinting methods, D/TGGE, T-RFLP, RFLP, SSCP are used to characterise of microbial community structure in soil (Nicomrat et al., 2006; Nicorarat et al., 2008; Sundberg et al., 2007; Truu et al., 2005). In addition, SSU rRNA clone libraries have provided fundamental information about the composition as well as the diversity of complex microbial communities. However, it is very complicated to reliably assess the number of microbial species, compare microbial communities and relate community composition to the environmental parameters in soil because of the estimation that one gram of soil can contain up to 10 billion of microorganisms from a thousand of different species (Torsvik and Ovreas, 2002). In order to assess full taxonomic diversity of microbes in environmental samples, high throughput DNA pyrosequencing (large 16S rDNA libraries) can be applied to obtain sufficient number of 16S rRNA gene sequences (Roesch et al., 2007). However, DNA based molecular techniques do not reveal information about the relationship between the identity and the function of microorganisms. Although metagenomic approaches and microarrays (GeoChip) have been developed in recent years for the direct monitoring of microbial genomic content in the environment, still only the genomic/degradation potential of the community can be described by these methods instead of realized activities (Truu et al., 2009). In order to get more than a functional prediction, gene transcripts (catabolic genes or the amount of 16S rRNA gene) and translated proteins must be obtained from environmental samples for the direct exa- mination. In addition, stable isotope probing (DNA and RNA-SIP in combination with the previously mentioned techniques) has been successfully applied in bioremediation studies enabling the linkage between microbial community structure and function (Uhlik et al., 2009).

(23)

23

4. MATERIAL AND METHODS 4.1. Site description

Phytoremediation and bioaugmentation field experiment was performed at the semi-coke depository area in Kohtla-Järve. The depository area consists of semi-coke mounds that have formed from solid waste of oil shale thermal treatment. These mounds have a shape of excentric cones, dark-gray or black in colour and with specific smell (Pae et al., 2005). Natural vegetation and under- growth are absent in the experimental area. In addition, there are no distinctive layers or horizons (including the humic layer) in the solid waste profile characteristic of the regular soils. According to Truu and coauthors (2003), semi-coke solid waste is characterized by a high initial pH value, a high concentration of sulphides, Ca²⁺ and Mg²⁺ ions and high amount of organic carbon (Table 2). The organic carbon in semi-coke is not similar to the organic carbon in regular soils because half of it is in the form of asphaltenic and bitumoid compounds which are very recalcitrant to biodegradation. Semi-coke has a granular texture, and the composition of the mineral part of semi-coke consist mainly of calcite, dolomite and ettringite (Motlep et al., 2007).

Test plots were established at the flat and older part (10–15 years) of the depository area in July 2001–2006 on the principle that the plots would have no influence on each other.

Table 2. Chemical properties of the solid waste at experimental area (Truu et al., 2003).

4.2. Phytoremediation experiment

In phytoremediation experiment, two different vegetation approaches were applied (Table 3):

I) inoculation of mixture of grass seeds onto the semi-coke (50 m² test plots).

The mixture of grass seeds was based on the four species: Lolium perenne – perennial ryegrass, Poa pratensis – Kentucky bluegrass, Festuca rubra – red fescue, and Festuca ovina – blue fescue. In addition to grass seeds, addi-

Variable Measured value

pH 8.0–11

Total nitrogen (%) 0.08

P-PO43– (mg kg^–1) 12.3

K⁺ (mg kg^–1) 799

Ca²⁺ (mg kg^–1) 18673

Mg²⁺ (mg kg^–1) 826

Total organic carbon (%) 15.0–18.0

Oil products (mg kg^–1) 340

Volatile phenols (mg kg^–1) 0.30–0.34

(24)

tional treatments were applied: sand treatment (grass seeds in semi-coke were covered by sand layer of 1–2 cm), peat treatment (grass seeds in semi- coke were covered by peat layer of 1–2 cm), no treatment (grass seeds in semi-coke). Part of the phytoremediation plots (10 m²) was used for bioaugmentation.

II) utilization of previously planted birches (Betula bendula; planted at the depository area in 1998; distance between birches is ca 1 m; 10 m² test plots) in semi-coke.

4.3. Bioaugmentation experiment

The previously characterized Pseudomonas strains isolated from the nearby polluted area were selected for the bioaugmentation experiment (characterization of the strains in detail in Heinaru et al., 2000; Merimaa et al., 2006; Table 3 in Reference III). The strains were Pseudomonas mendocina (PC1), P. fluore- scens biotype F (PC17 and PC20), P. fluorescens biotype B (PC18) and P.

fluorescens biotype C (PC24). Microbial strains used in current study are de- posited in the Collection of Environmental and Laboratory Strains of Tartu University (CELMS, http://www.miccol.ut.ee). Various mixtures of Pseudo- monas strains were applied simultaneously with vegetation (Table 3):

I) in combination with grass species, a mixture of three bacterial strains (PC1, PC24 and PC18; strains in ratio 3:1:1, respectively).

II) in combination with birches, a) a mixture of four bacterial strains (PC1, PC18, PC20 and PC24; strains in equal ratios), and b) a mixture of five bacterial strains (PC1, PC17, PC18, PC20 and PC24; strains in equal ratios).

Before the introduction, the strains were cultivated in Luria–Bertani (LB) medium. Cells from the stationary growth phase were mixed with 0.9% NaCl solution and inoculated onto the surface of experimental plots. Inoculation was performed by spreading the mixture (20 L; final concentration of each strain approximately 10⁸ CFU ml^–1; the total amount of bacteria introduced into the plots was ca 10¹² CFU m^–2) onto the surface of experimental plots. Inoculation was performed in July 2002–2006 only to the plots established in the same year.

4.4. Chemical analyses

Oil products (extracted with pentane) were measured by gas chromatography.

Volatile phenols were determined spectrophotometrically. Total orgacic carbon was determined using an infrared spectrophotometer. All analyses were carried out by Tartu Environmental Research, Ltd. (Tartu, Estonia).

(25)

25

Table 3. Principal scheme of the phytoremediation and bioaugmentation field experi- ments at the semi-coke depository area.

Phytoremediation grass seeds previously planted birches Bioaugmentation PC1, PC18,

PC24 PC1, PC18, PC20, PC24

PC1, PC17, PC18, PC20, PC24

Plots establishment 2001 2004 2006 2005

Bacterial biomass

inoculation in July 2002 2004 2006 2005

Additional treatments no treatment, sand layer; peat

layer – – mineral

fertilizers Soil sampling in October 2001–2003 2004–2007 2006–2007 2005–2007

4.5. Soil sampling and bacteria cultivation dependent analyses

Soil sampling was performed during 2001–2007 at the beginning of October.

Bulk soil and rhizosphere sampling was performed as described in Reference III. Bulk soil and rhizosphere samples were used for serial dilutions and plating selective media. The numbers of aerobic heterotrophic and phenol degrading bacteria were determined as described in References I and II. In addition, obtained bacterial colonies from selective media were characterized by different analyses (Gram test, oxidase test, growth on King’s B medium, presence of catabolic genes, rep-PCR) as described in Reference III for the detection of introduced strains from semi-coke.

Potential metabolic activity of the microbial communities in bulk soil and rhizosphere samples were characterized using Biolog EcoPlates and substrate- induced soil respiration tests as described in References I and III, respectively.

Substrate utilization dynamics of microbial communities obtained from Biolog EcoPlates were used for Principal Component Analyses (PCA) to assess the changes in culturable microbial community composition due to vegetation and bioaugmentation.

4.6. Molecular analyses

Microbial DNA was extracted from the samples using UltraClean Soil DNA kit (Mo Bio Laboratories) according to the manufacturer´s instructions. In addition, protocol by Peršoh and coworkers (2008) was used for the isolation of total nucleic acids (DNA and RNA) from the samples (description of the isolation and subsequent analyses are given below).

7

(26)

Microbial communities in semi-coke were characterized based on the sequences of total bacterial 16S rRNA gene and 16S rRNA specific cDNA, genus Pseudomonas specific 16S rRNA gene and 16S rRNA specific cDNA, and large subunit of multicomponent phenol hydroxylase gene (lmPH). PCR amplification was performed, using following primers: PRBA338f, PRUN518r, PRBA338f-GC for universal bacterial 16S rRNA gene and 16S rRNA specific cDNA; pseF, pseR, pseF2 (5´-GGTCTTCGGATTGTAAGCAC-3´) for genus Pseudomonas specific 16S rRNA gene and 16S rRNA specific cDNA; phe00, phe212, pheGC for lmPH gene (nucleotide sequences of used primers are given in tabel 4 in Reference III). Amplification of total bacterial 16S rRNA gene is described in Reference II; amplification of genus Pseudomonas specific 16S rRNA gene is described in Reference III; amplification of lmPH gene fragments is described in Reference III.

Amplified products were applied for DGGE analyses, resulting in the fingerprints of the samples. Comparison of the fingerprints were perfomed to assess the changes in the microbial community composition or the diversity of catabolic genes caused by vegetation and bacterial strains as well as detection of the gene fragments corresponding to the introduced bacterial strains. Similarity values based on densitometric curves of the gel tracks were calculated using the Pearson correlation coefficient. Dendrogram based on cluster analyses of the DGGE profiles was performed. Also PCA was used for the analysis of DGGE banding patterns. DGGE analyses was performed as described in Referencess II and III, subsequent cloning and nucleotide sequencing was performed as described in Reference III in order to verify, whether the detected 16S rRNA ja lmPH gene fragments from soil samples show similarities to sequences from the introduced bacterial strains.

4.6.1. Isolation of total nucleic acids from the soil and subsequent analyses

Protocol by Peršoh et al. (2008) was used for the isolation of nucleic acids from the soil samples in order to assess the impact of vegetation and introduced bacterial strains on the dynamics of microbial community composition in the levels of both, DNA and RNA. Nucleic acids isolation was performed in three replicates per sample. Obtained DNA was then used for the DGGE and quantitative PCR (QPCR), targeting the total bacterial and genus Pseudomonas specific 16S rRNA gene (Fig. 3). Part of the obtained samples from nucleic acids extraction were treated with the Dnase I (Fermentas), followed by the cDNA synthesis (RevertAid M-MuLV Reverse Transcriptase; Fermentas) according to the protocols provided by the manufacturer. cDNA synthesis was performed with 16S rRNA gene specific primer: primer PRBA338f for the total bacterial 16S rRNA gene, and primer PseF for the genus Pseudomonas specific 16S rRNA gene. The obtained total bacterial 16S rRNA specific cDNA was

(27)

27

used for QPCR and DGGE, obtained genus Pseudomonas 16S rRNA specific cDNA was used for DGGE (Fig. 3).

QPCR was performed using Platinum SYBR Green qPCR Supermix-UDG kit (Invitrogen) according to the manufacturer´s instructions. Universal bacterial primers PRBA338f and PRUN518r were used for the total bacterial 16S rRNA gene and 16S rRNA specific cDNA, primers PseF2 and PseR were used for the genus Pseudomonas specific 16S rRNA gene and 16S rRNA specific cDNA.

The reaction conditions for the QPCR were as follows: 50ºC for 2 min, 95ºC for 2 min, 40 cycles of 95ºC for 15 s, 60ºC for 30 s, 72ºC for 30 s, followed by melting curve analysis to confirm the fluorescence signal resulted from specific PCR products and not primer-dimers or other artifacts. QPCR was performed in triplicate for each sample, including standards and negative control. Reactions were carried out with an ABI Prism 7900HT machine (Applied Biosystems, Foster City, CA, USA) and data were analysed using S.D.S 2.2.2 software (Applied Biosystems). In addition, the abundance of Pseudomonas sp. group was calculated as the ratio between the measured copy numbers for Pseudo- monas sp. specific 16S rRNA gene and cDNA of 16S rRNA, and the total bacterial 16S rRNA gene and cDNA of 16S rRNA, respectively.

For standard curves, both targeted sequences were amplified from positive control strain Pseudomonas mendocina PC1. Amplified products were run on 2% agarose gel to confirm the specificity of the amplification, and cloned into vector pTZ57R using InsT/Aclone^TM PCR Product Cloning kit (Fermentas).

Plasmids were isolated using QIAprep Spin Miniprep kit (Qiagen, CA, USA).

In case of using DNA as template, plasmid DNA concentration were determined with spectrophotometer (Nanodrop 1000) and standard curves were obtained with serial plasmid dilutions of a known amount of plasmid DNA containing the targeted fragment. In case of using cDNA as template, the plasmids were linearized with EcoRI and purified with UltraClean 15 DNA Purification Kit From Gels and Solutions (MoBio Laboratories Inc) followed by in vitro transcription with T7 RNA polymerase (Fermentas) according to the protocol provided by manufacturer. Obtained sample was treated with the Dnase I (Fermentas) according to the protocol and purified with UltraClean 15 DNA Purification Kit From Gels and Solutions kit (MoBio Laboratories Inc).

RNA concentration was determined spectrophotometrically (Nanodrop 1000).

Ten-fold serial dilutions (20 pg μl^–1–0.0002 pg μl^–1) of the RNA was used for the cDNA synthesis using total bacterial 16S rRNA gene specific primer PRBA338f according to the protocol (Fermentas) followed by the QPCR.

Contamination with DNA was tested with the control PCR. Copy numbers were calculated from the standard curves, assuming that the average molecular mass of a double-stranded DNA molecule is 660 g mol^–1. Copy numbers of the samples were quantified by comparing the cycle at which fluorescence crossed a threshold to a standard curve.

DGGE fingerprinting of both, the total bacterial and genus Pseudomonas specific 16S rRNA gene and 16S rRNA specific cDNA was performed. Total bacterial 16S rRNA gene was first amplified with primers PRBA338f and

(28)

PRUN518r followed by amplification with primers PRBA338fGC and PRUN518r, while the genus Pseudomonas specific 16S rRNA gene was first amplified with primers PseF and PseR followed by amplification with primers PRBA338fGC and PRUN518r (Fig. 3). Total bacterial and genus Pseudomonas specific 16S rRNA cDNA were amplified with primers PRBA338fGC and PRUN518r (Fig. 3). Primer PRBA338fGC contains a GC clamp (40 bp) at the end of 5´ end to enable DGGE analyses. The PCR mixture included 1 × PCR buffer [with (NH4)2SO4], 200 μM concentrations of each dNTP, 2.5 mM MgCl2, 20 pmol of both primers, 60 ng μl^–1 of bovine serum albumin and 0.5 U of Taq DNA polymerase (Fermentas). The reaction conditions of the PCR were as follows: 95°C for 5 min, 30 cycles of 95°C for 1 min, 57°C for 1 min, 72°C for 1.5 min, and a final extension of 72°C for 5 min. The PCR product were quantified in 2% (w/v) agarose gel by comparison with the standard (100 bp DNA size marker, Fermentas) using EASY WIN32 software (Herolab GmbH).

Approximately 500 ng of the PCR products were applied for the DGGE analysis as described in Reference III. A linear denaturing gradient of 35–62%

was used for the total bacterial community, and 38–49% for the genus Pseudo- monas community. Subsequent cloning and nucleotide sequencing of the frag- ments specific from genus Pseudomonas community was performed as described in Reference III.

Figure 3. Schematic representation of steps of analyses followed by DNA and RNA based QPCR and DGGE. Nucleotide sequences of the used primers are given in table 4 in Reference III.

(29)

29

5. RESULTS AND DISCUSSION

5.1. Impact of vegetation and bioaugmentation on the concentration of organic pollutants

in semi-coke

Chemical analyses were performed to assess the impact of vegetation and bioaugmentation on the concentration of different pollutants in semi-coke. As mentioned earlier, semi-coke consists of various contaminants some of which are recalcitrant to biodegradation (PAHs and asphaltenic compounds for example). However, chemical analyses revealed that vegetation and introduced bacterial strains had substantial impact on the removal of the contaminants.

Within a two and a half year period the concentration of oil products decreased more than three times in the plots with vegetation and added bacterial biomass (from 340 mg kg^–1 to 100 mg kg^–1, Reference I). In addition, the concentration of volatile phenols reduced up to 100% and the total content of organic carbon decreased by 10 to 30 g kg^–1 (from 15% to 12%) on average in plots with grass species compared to the control plot without vegetation (Reference I). In case of phenols removal it is known, that in addition to the biodegradation, phenolic compounds are easily to be removed from the soil by leaching. In case of some other pollutants adsorption of the contaminant to the soil particles may occur reducing the mobility and bioavailability of the contaminant instead of biodegradation of the pollutant. However, in our experiment, the reduction of the concentrations of oil products and phenols was found to be exceptionally high in upper soil layer with the highest root density referring to the biodegradation by microorganisms. It is generally recognized that enhanced biodegradation activity in the rhizosphere is due to rhizodeposition consisting of root exudates and root debris. Also, release of glutathione conjugates into the rhizosphere by plants during detoxification process has been shown, where they could be metabolized by microorganisms (Schroder et al., 2007). Since the degradation rates of pollutants did not differ significantly between plots with vegetation we may suggest that establishment of vegetation on semi-coke was the key factor for the acceleration of pollutants degradation (Reference I).

5.2. Impact of vegetation and bioaugmentation on the microbial abundance

Different mechanisms for the facilitation of phytoremediation have been shown for different plant species. Kirk and coauthors (2005) found, that perennial ryegrass supported general increase in microbial activity and numbers in the rhizosphere, some of which had catabolic activity towards petroleum hydrocarbons in petroleum-contaminated soil, while alfalfa specifically increased the number of microorganisms capable of degrading more complex hydrocarbons.

8

(30)

We also studied the impact of vegetation and bioaugmentation on the microbial abundace. Results from earlier study showed, that vegetation increased the number of biodegradative bacteria as well as the activity and diversity of microbial community in semi-coke, while the number of aerobic heterotrophic bacteria remained at the same level compared to the plot without vegetation (Truu et al., 2003). However, these results were obtained only four months after the inoculation of the grass seeds into the semi-coke. By assessing microbial counts over longer time period we may suggest, that neither vegetation nor introduced bacterial strains had influence on the number of aerobic heterotrophic bacteria (AHB) in semi-coke, but had short term effect on the number of phenol degrading bacteria (PDB). The numbers of AHB were in the range of 10⁶–10⁷ CFU g^–1 soil (dw) in bulk soil of both planted samples, unplanted control, and bioaugmented samples (Table 2 in Reference II; Table 4). The numbers of PDB were one order of magnitude higher in bulk soil samples from plots with vegetation and added bacterial biomass on the first year of the experiment (four months after the addition of bacterial strains to the semi-coke) compared to the unplanted/planted control (ca 10⁵ and 10⁴ CFU g^–1 soil dw, respectively), but these numbers decreased to the level of control sample next year (Table 2 in Reference II; Table 4). The dynamics of bacterial counts were similar between two vegetation approaches. Rhizosphere samples from different plots demonstrated generally one to two orders of magnitude higher numbers of both bacteria compared to the numbers obtained in bulk soil and these numbers remained stable during the experiments. Similar dynamics could be obtain when assessing the proportion of PDB in total bacterial abundance during the experiment – the proportion of PDB in bulk soil from plots with birches and added bacterial strains decreased within four months of bioaugmentation to the level of 20%, and to the level of control sample in following months (Table 4).

The proportion of PDB in rhizosphere of birches followed similar dynamics as in bulk soil, although in smaller scale.

The problem with these results is that bacteria enumeration depends largely on the ability of bacteria to grow at the laboratory conditions. To obtain cultivation independent data about the abundance of bacteria, isolation of nucleic acids directly from semi-coke was performed for the quantitative PCR.

Estimation of the copy numbers of total bacteria and genus Pseudomonas specific 16S rRNA gene (DNA based analyses) and cDNA of 16S rRNA (16S rRNA gene transcript – RNA based analyses) was performed. 16S rRNA gene as commonly used marker is present in all bacterial species, and 16S rRNA gene transcript is even more abundant in RNA extracts because its high transcription rate in metabolically active cells.

(31)

Table 4. The impact of birches and added bacterial strains (combination of four strains) on the numbers of aerobic heterotrophic bacteria (AHB) and phenol degrading bacteria (PDB), and on the copy numbers of total bacterial and genus Pseudomonas specific 16S rRNA gene and cDNA of 16S rRNA. Mean values of copy numbers are presented (n = 9; stdev < 15% of the mean value). SampleAHB (CFU g–1 soil dw) PDB (CFU g–1 soil dw) PDB/AHB 16S rRNA gene cDNA of 16S rRNA16S rRNA gene copy ratio Psea /total

16S rRNA cDNA copy ratio Psea /total

total bacteria genus Psea total bacteriagenus Psea No. copy g–1 soil dw R4 8.22E+065.19E+06 0.63 9.47E+087.92E+08 1.04E+07 6.21E+060.83650.5971 BC 1.58E+063.49E+04 0.02 7.64E+061.37E+05 2.29E+07 1.06E+050.01790.0046 BC* 1.58E+071.46E+05 0.01 1.69E+091.12E+07 6.79E+10 1.72E+090.00660.0253 4 1.37E+062.75E+05 0.20 5.60E+061.49E+05 7.47E+07 5.23E+050.02650.0070 4* 1.24E+078.57E+05 0.07 2.45E+081.59E+06 3.85E+10 1.12E+090.00650.0291 16 2.50E+061.60E+04 0.01 2.16E+061.86E+05 2.58E+07 1.32E+050.08590.0051 16* 2.09E+075.86E+05 0.03 7.68E+081.00E+06 4.53E+10 1.07E+090.00130.0236 28 3.03E+062.22E+04 0.01 6.41E+066.28E+05 1.20E+07 8.47E+040.09790.0071 28* 1.92E+073.65E+05 0.02 4.65E+081.85E+06 6.37E+10 1.33E+090.00400.0209 AHB – aerobic heterotrophic bacteria; PDB – phenol degrading bacteria; R4 – soil sample was obtained 2h after the introduction of the bacterial strains onto the semi-coke; BC – control with birches; * – rhizosphere sample; sample numbers indicate bioaugmentation age in months;a – Pseudomonas sp.

JAANIS JUHANSON Impact of phytoremediation and bioaugmentation on the microbial

JAANIS JUHANSON Impact of phytoremediation and bioaugmentation on the microbial

community in oil shale chemical

industry solid waste

CONTENTS

LIST OF ORIGINAL PUBLICATIONS

ABBREVIATIONS

1. INTRODUCTION

2. THE AIM OF THE STUDY

3. LITERATURE REVIEW 3.1. Phytoremediation

3.2. Bioaugmentation

3.3. Microbial community structure in the vicinity of roots

3.4. Monitoring bioremediation

4. MATERIAL AND METHODS 4.1. Site description

4.2. Phytoremediation experiment

4.3. Bioaugmentation experiment

4.4. Chemical analyses

4.5. Soil sampling and bacteria cultivation dependent analyses

4.6. Molecular analyses

5. RESULTS AND DISCUSSION

5.1. Impact of vegetation and bioaugmentation on the concentration of organic pollutants

in semi-coke

5.2. Impact of vegetation and bioaugmentation on the microbial abundance